how to set up livekit observability

Thanks for reaching out, Jupiter. I'm an AI assistant and I'm here to help. Give me a moment while I research this topic.

I use livekit cloud. yeah I mean agent observability

Great — since you’re on LiveKit Cloud and want Agent Observability, here’s how to enable and use it. 1. Enable observability for your project • Go to your project’s Settings in the LiveKit Cloud dashboard: https://cloud.livekit.io/ • Open the Data and privacy section. • Toggle Agent Observability on at the project level (it’s off by default). 1. Use the Agent insights views • Open the project → Sessions → select a session → Agent insights tab. • Views available: ◦ Transcript: conversation timeline with audio playback. ◦ Trace: turn-by-turn timing for LLM, TTS, end-of-turn detection, and tool calls. ◦ Logs: chronological info/warn/error/debug messages across the stack. 1. Per-agent control (optional) • To disable recording/observability for a specific agent, set

record=False

when creating the AgentSession:

session = AgentSession(..., record=False)

1. Collect metrics programmatically (optional) • Use the

UsageCollector

to capture real-time metrics (TTFT, TTFB, token usage, turn latency):

from livekit.agents import metrics

usage_collector = metrics.UsageCollector()
session.on("metrics_collected", lambda ev: usage_collector.collect(ev.metrics))

1. Download and retention • You can download audio, transcripts, and logs from a session. • Observability data is retained for 30 days (stored in the US) and is free during the beta period through end of 2025 (monthly plan allotments apply). If you’d like, I can: • Walk you through the exact clicks in your dashboard while you’re there, or • Provide sample code for enabling/disabling recording in your agent code, or • Show how to interpret Trace metrics for diagnosing interruptions. Which of those would you like next?